Extreme Element Enrichment by the Interaction of Supercritical Fluids from the Mantle with Crustal Rocks

Abstract

We show that supercritical fluids or melts are verifiable by critical high-temperature and high-pressure minerals like diamond, lonsdaleite, and others in crustal rocks as foreign minerals. In combination with the pseudo-binary solvus curves (temperature versus water content of silicate melts) with the Lorentzian distribution of some ore-forming elements, which are untypical for hydrothermal forming processes, we have solid proof for the interaction of mantle fluids and crustal rocks (granites). In this contribution, we restrict ourselves to a small number of critical observations, in particular on cassiterite polytypes.

1. Introduction

Since the development of the method for the determination of water in glass and silicate melt inclusions [1], the further improvement of this method [2,3], and the completion of the determination of boric acid [H₃BO₃] [4], we can show that many pegmatites form a pseudo-binary solvus curve in the coordinates water concentration versus temperature. Beginning with the work on the determination of beryllium in pegmatitic melt inclusions [5], it has become more evident from year to year that the solvus curves are also regularly related to trace and some main elements. The maximum of elements are clearly associated with the solvus crest, forming Gaussian or Lorentzian element distribution patterns [6,7]. Figure 1 and Figure 2 exemplarily show such solvus curves and the corresponding Lorentzian distribution of tin. The connection between solvus and, e.g., the Lorentzian distribution is in [8], presented as a three-dimensional plot outlined.

Table 1. Lorentzian fit parameters for the Sn distribution.

Area	Center	Width	Offset	Height	R2
51,522	25.6% H2O	4.77% H2O	504 ppm Sn	6882 ppm	0.9455

provides the parameters for the Lorentzian distribution of Sn (ppm) for Ehrenfriedersdorf pegmatite.

The Clarke number of Sn for granitic rocks is three ppm [9]. Tin granites of the Erzgebirge have a value of about 200 ppm [10]). The offset is already 2.5-fold higher than the Sn value of highly specialized tin granites.

The relationship between solvus and Lorentzian distribution of elements, here Sn, is typical for pegmatites, and we interpret that as the first hint for the interaction of supercritical fluids with granites and other crustal rocks. At the solvus crest (critical point, CP), we observe the highest element enrichment, here about 0.6 (mol/mol) SnO₂. Further studies confirmed that the highest measured Sn value was 16,430 ppm at 26.5 (g/g) H₂O, corresponding to 1.39 (mol/mol) SnO₂. That means, at least, that the Lorentzian distribution is a first approximation. Much work is necessary to solve these complex questions. Such distributions (pseudo-binary solvus curves (H₂O-silicate melt versus temperature and Lorentzian element distribution)) are untypical for processes in the normal hydrothermal state, which is far away from the supercritical state here considered. High values for tin solubility were obtained by [11], with hydrothermal diamond anvil cells and by using alkali carbonate aqueous solutions. They received for the Li₂CO₃-H₂O system 0.41 (mol/mol) SnO₂ at about 790 °C and 795 MPa. These values are much lower than the supercritical conditions; however, they are not far from the supercritical state.

Studies on alkali chlorides and carbonates in diamonds [12,13] have resulted in further studies on hints to the interaction of supercritical fluids and crustal rocks. Experimental and theoretical works on the problem of supercritical fluids are in [11,14,15]. The historical development of research on supercritical fluids has been given by [16].

Studies on the Rønne pegmatite (Bornholm Island) yield clear proof that alkali carbonate-rich (NaHCO₃ Li₂CO₃ and others) supercritical fluids, together with ¹³C-rich diamond, take part in the formation of pegmatites [17].

Starting with new research on the prismatine rock from Waldheim/Saxony [18], the first author found clear indications that supercritical fluids or melts have influenced the well-known metamorphic rock. Hints for that are slight spheric and total smooth, polished-like mineral inclusions such as diamond, reidite, coesite, and stishovite. These findings initiate the search for such smooth spherical mineral inclusions in crustal rocks. The first object was granite from Greifenstein, near Ehrenfriedersdorf, Central Erzgebirge, Germany.

2. Sample Materials and Methodology

2.1. Sample Materials

The sample material used here is a tetragonal yellow-colored cassiterite crystal (sample Sn 40 from the Mining Academie Freiberg, cited in [19], 1 cm in diameter from Krupka (Krušnohoří Mining District/Czech Republic). All other sample material is mentioned in the text and cited in the corresponding references. More information on the Krupka deposit is in [20,21,22].

2.2. Methodology

For the preparation of the presented results, we used only microscopic and Raman spectroscopic techniques.

2.3. Raman Spectroscopy

We performed all microscopic and Raman spectroscopic studies with a petrographic polarization microscope with a rotating stage coupled with the EnSpectr Raman spectrometer R532 (Enhanced Spectrometry, Inc., Mountain View, CA, USA) in reflection and transmission. The Raman spectra were recorded in the spectral range of 0–4000 cm⁻¹ using an up-to-50 mW single-mode 532 nm laser, an entrance aperture of 20 μm, a holographic grating of 1800 g/mm, and spectral resolution ranging from 4 to 6 cm⁻¹. Generally, we used an objective lens with a magnification of 100×—the Olympus long-distance LMPLFLN100x objective (Olympus, Tokyo, Japan). The laser power on the sample is adjustable down to 0.02 mW. The Raman band positions were calibrated before and after each series of measurements using the Si band of a semiconductor-grade silicon single-crystal. The run-to-run repeatability of the line position (based on 20 measurements each) is ±0.3 cm⁻¹ for Si (520.4 ± 0.3 cm⁻¹) and 0.4 cm⁻¹ for diamond (1332.7 ± 0.4 cm⁻¹ over the range of 80–2000 cm⁻¹). The FWHM = 4.26 ± 0.42 cm⁻¹. FWHM is the Full-Width at Half Maximum. We also used a water-clear natural diamond crystal (Mining Academy Freiberg: 2453/37 from Brazil) as a diamond reference (for more information, see [23,24]).

3. Results

The solvus curves and the Lorentzian element distribution patterns (up to now for about 20 element and molecule groups derived [6,7]) alone are not enough to say that the supercritical fluids come from mantle depths. At high temperatures, where water-rich pegmatite melts form, the supercritical state can create and work. Therefore, further proof is necessary. This comes from the evidence of mantle minerals in ore deposit minerals trapped at low-pressure crustal regions.

In the Greifenstein granite (1–2 kbar: shallow intrusion level [25]), we found spherical minerals like high-pressure cristobalite, coesite, diamond, lonsdaleite, kumdykolite, moissanite, high-pressure beryl, zircon, zircon-reidite, and OH-dominated topaz [24]). Because such minerals and their unique form cannot be formed at the crustal range, we interpret that as hints of the impact of supercritical fluids or melts coming from the mantle region. The first signs of the participation of mantle components come from [26]. These authors found, however, no compelling pieces of evidence for their statement.

The present author, together with some coauthors, has shown that the supercritical fluids/melts bring together untypical minerals (diamond, moissanite, and many others), especially high-temperature water, CO₂, and CH₄, into the crustal region. Thus, we often found orthorhombic cassiterite, diamond, and graphite inclusion in the Variscan cassiterites of the German Erzgebirge, Krušnohoří, and Slavkovský les (Czech Republic) regions. In 18 cassiterite samples from there, we could show that inclusions of orthorhombic cassiterite (CaCl₂ polytype Pnnm, PbO₂-polytype Pbcn, ZrO₂-polytype Pbca, and cotunnite-polytype Pnma-I) are widespread (for the SnO₂-polytypes, see [27]. The Pnma-I (cotunnite-type of cassiterite) is from other orthorhombic polytypes that are relatively rare. There is also often graphite, diamond, and sometimes moissanite primarily present. In the sample Sn-40, we could also prove complex C-B-N carbide phases (see [28]. Figure 3a–e show such inclusion of orthorhombic cassiterite in normal tetragonal, pale-yellow-colored cassiterite (Sn 40, in [19]) from Krupka (Krušnohoří Mining District/Czech Republic).

Figure 3b shows a typical Raman spectrum of a tetragonal cassiterite host (Sn-40) with the typical Raman bands at 632 and 775 cm⁻¹.

The cotunnite-type cassiterite (Pnma-I) shows the Raman main band at significantly lower values (626 versus 632 cm⁻¹).

Figure 3d represents the Raman bands of the cotunnite-type cassiterite (Pnma-I) in the low-frequency range.

To completion, the Raman spectrum Figure 3e shows the two typical OH bands ν₂ bending vibration (H-O-H) at 1628 and ν₁ at 3408 cm⁻¹.

Table 2. Raman shift (cm⁻¹) of an orthorhombic cassiterite inclusion in yellow tetragonal cassiterite from Krupka (Krušnohoří Mining District/Czech Republic).

Orthorhombic Cassiterite (Pnma-I) as Inclusion in Cassiterite from Krupka
Raman Band	Raman Mode	Relative Intensity
138.2 ± 2.2	Ag	vs
159.5 ± 2.6	B1g	vs
232.0 ± 2.4	Ag	s
474.0 ± 4.8	B1g	w
625.8 ± 2.0	Ag	vs
691.3 ± 0.6	B1g	vw

Remark: The Raman lines are not stable under the 532 nm laser (50 mW). The best data can be obtained with a low laser power of 0.3 mW on samples and a long measuring time of 2000 s. Here is a change in the Pbcn-cassiterite polytype. vs—very strong; s—strong; w—weak; vw—very weak.

shows the Raman main lines of the orthorhombic cassiterite (Pnma-I)—from the cotunnite polytype of cassiterite (see [27])—10 measurements.

The cotunnite-type cassiterite inclusion also contains remnants of other orthorhombic cassiterite polymorphs, such as the CaCl₂-type cassiterite (see [31]). According to [27,30]), the formation of the cotunnite-type polymorph of cassiterite happens at about 19.7 GPa and 1200 K.

During the transition from a supercritical state to a critical and under-critical state, chemical and physical processes happen, which cannot be solved at the moment. As an example, the formation of moissanite whiskers in beryl in the Variscan tin deposit Ehrenfriedersdorf, Sauberg mine, can be referenced [32]. The growth of moissanite at temperatures lower than 700 °C is untypical.

At this point, we remember that the fundamental physical properties of supercritical fluids, like viscosity and diffusivity, must go in the direction of zero and infinite, respectively (e.g., [14]). This basic behavior of supercritical fluids determines all other properties of such fluid—ascent velocity, solubility behavior, complex formation, surface tension, and many others (see [8]).

Furthermore, ref. [20] has clearly shown with his sketches (Figure D in this contribution) that the amount and distribution of small and large pegmatite bodies in the Sauberg mine near Ehrenfriedersdorf cannot be formed by the water of the surrounding granites. A different, intense source of water is necessary. Starting with the interpretation of the origin of mantle minerals in the prismatine granulite von Waldheim/Saxony, it would be more evident that supercritical fluids have had a more significant impact as of now, though. The question of where the ore elements come from immediately follows. Is the classic idea correct that, for example, Sn comes from the so-called surrounding tin granite alone? Or do we have a different or second source? And is the high level of 200 ppm Sn in specialized granites not the result of the interaction of supercritical fluids with granites in the upper crust itself?

Scientific proof [31] clearly shows that in many tin deposits of the German Erzgebirge and from Krušnohoří Mining District, Czech Republic, the Sn comes as high-pressure and high-temperature orthorhombic cassiterite together with supercritical fluids, also bearing solved tin and other elements from a depth corresponding to a pressure of 15 GPa or more, which demands a reassessment of the conventional wisdom up to now.

4. Discussion

Our careful study of samples from the tin deposits of the Erzgebirge Mts., Germany, shows that many minerals (quartz, topaz, and cassiterite) contain high-temperature and high-pressure minerals, which is unmistakable proof of the interaction between mantle and crust by supercritical fluids or melts.

Figure 4 shows such a diamond crystal in cassiterite and the corresponding Raman spectrum. Often, we find in cassiterite crystals, besides the rare diamond crystals from the Variscan Erzgebirge and the Slavkovský les, North Bohemia (Czech Republic), orthorhombic cassiterite inclusions in tetragonal cassiterite [31], demonstrating that at least a part of the cassiterite comes from mantle depths indicated by the high-pressure (~15 GPa or more) formation of orthorhombic cassiterite.

Figure 5 shows an orthorhombic (Pbcn polytype) cassiterite crystal in the tetragonal cassiterite matrix (Sn-40) from Krupka with diamond and graphite inclusions. Most orthorhombic cassiterites show replacement structures by the tetragonal cassiterite. Some cassiterite crystals contain up to 20 [%(vol/vol)] orthorhombic cassiterite phases (e.g., Sn-49).

The observation of high-pressure and high-temperature minerals is not only related to the Variscan Erzgebirge and the Krušnohoří Mining District, Czech Republic region. We found such proof in the Königshain granite, in some quartz veins of the Lusatian Mts, and also in East Thuringia [33]. During the newest Raman studies on the pegmatites related to the Rønne granite (Bornholm Island), the first author found much ¹³C-rich diamond and ¹³C-rich graphite and other surprising findings like NaDCO₃^-rich phases and ¹³C-rich CO₂ [17].

All these results obtained in the last three years, together with the coauthors’ findings, show that the meaning of supercritical fluids or melts in the case of Variscan tin deposits was not taken into account. Obviously, supercritical fluids play an essential role in the formation of pegmatites and other high-temperature ore deposits. Besides critical high-temperature and high-pressure minerals like diamond and lonsdaleite, we also found inclusions of ore-forming minerals (different cassiterite polytypes, caracolite [Na₃Pb₂(SO₄)₃Cl], REE-phosphates, and many others; see [24]). Furthermore, the presence of inclusions or remnants of high-pressure and high-temperature SnO₂ polymorphs of rutile-type → CaCl₂-type → pyrite-type → ZrO₂ orthorhombic phase I → cotunnite-type [27,30] demonstrate that high-pressure phases (CaCl₂- and cotunnite-type) are essential pieces of evidence for the transport of this ore mineral from mantle depths to the crust region. This observation shows further clear indications that the element solubility in supercritical phases can have extreme values, and the concentration of tin indicated in the form of a Lorentzian curve with the maximum at the solvus crest is not the highest Sn concentration in the supercritical fluid. This concentration reflects the transition from the supercritical to critical state.

This results in the following question: Are the typical fluid inclusions in cassiterite only (e.g., [19]) proof of the recrystallization and re-deposition of cassiterite under hydrothermal conditions? Up to now, the tin-rich melt inclusions in cassiterite or the high-pressure and high-temperature polymorphs of cassiterite inclusions do not feature in genetic discussions. Is there an analog situation like the emerald formation via supercritical fluids in the Habachtal, Austria [34], where the dominant fluid inclusions represent only the secondary processes and the up-to-now ignored melt inclusions are the primary inclusions generated by supercritical fluids? By this acceptance of supercritical fluids, the contradiction between regional geology and the observation of high-temperature indications can be solved. The proof of diamond (1329.2 cm⁻¹ with an FWHM = 76.3 cm⁻¹) by the first author in the Habachtal emerald supports this statement.

Compared to diamond and lonsdaleite inclusions, which are indicative of mantle origin, different high-pressure and high-temperature cassiterite polymorphs are pressure and temperature sensors for the origin of supercritical fluids.